tls: do not wrap net.Socket with StreamWrap

[krydos]

Hi :)
In this PR I just wanted to avoid "c-like" error from this issue - #3655. You can see below how error looks like now. At least there is stacktrace exist.

Please update me if I did something horribly wrong here or maybe if this is not a fix at all so I will close this PR.

_stream_wrap.js:42
  this.stream.pause();
              ^

TypeError: this.stream.pause is not a function
    at new StreamWrap (_stream_wrap.js:42:15)
    at new TLSSocket (_tls_wrap.js:280:12)
    at Object.<anonymous> (/path/to/test-file.js:4:1)
    at Module._compile (module.js:582:30)
    at Object.Module._extensions..js (module.js:593:10)
    at Module.load (module.js:516:32)
    at tryModuleLoad (module.js:466:12)
    at Function.Module._load (module.js:458:3)
    at Function.Module.runMain (module.js:618:10)
    at startup (bootstrap_node.js:144:16)

Checklist

• make -j4 test (UNIX), or vcbuild test (Windows) passes
• tests and/or benchmarks are included
• commit message follows commit guidelines

Affected core subsystem(s)

[mscdex]

Can you include a test that fails without this change?

[krydos]

test added. Thank you @mscdex .

test expect TypeError and do not check the exact error message. I did it like that mostly because test should check the fact that exception was thrown.

[mscdex]

/cc @nodejs/crypto

[addaleax]

Looks good to me, but to be careful we should maybe consider this semver-major (@nodejs/ctc), and it would be good to have @indutny give this a look if possible.

[addaleax]

Also, can you edit the commit message to add a Fixes: https://github.com/nodejs/node/issues/3655 line? :)

[krydos]

@addaleax I've update commit message, thank you. And actually thank you for helping me to understand and fix the issue :)

[addaleax]

Labelled semver-major to be on the safe side

CI: https://ci.nodejs.org/job/node-test-commit/9635/

[krydos]

I think CI hates me 😢

Could you please approve I understand CI output correctly, and it is not related?

[lpinca]

@krydos gonna start a new one.

CI 2: https://ci.nodejs.org/job/node-test-pull-request/7880/

[cjihrig]

I know you commented on this already, but please check the full error message.

[indutny]

LGTM

[indutny]

Minor nit, lots of unnecessary parens. Do we allow them with lint?

[krydos]

linter was quiet in this case. I'm also wasn't able to find rule to check this...

[krydos]

@cjihrig I've added full error message to the test.
@indutny unnecessary parantecies are removed.

Thank you for the feedback!

[addaleax]

I think @indutny was just talking about the inner parentheses – I would really prefer to keep the ones around the && expression for readability

[krydos]

@addaleax oops, didn't understand it correctly. Sorry.
I've added parentheses around && expression. Thank you :)

addressed

[addaleax]

Landed in b23d414

[addaleax]

Looking back, I don’t think labelling this semver-major is necessary – I’d prefer to wait a bit before backporting, though.

[krydos]

@addaleax agree with you. Looks more as a bugfix. I hope there are no users of Node.js who rely on this error.

[gibfahn]

I hope there are no users of Node.js who rely on this error.

CitGM run: https://ci.nodejs.org/view/Node.js-citgm/job/citgm-smoker/796/

[addaleax]

@gibfahn The “error” in this case is a failed C++ assertion (i.e. hard crash), just in case you were thinking this is about error messages. :)

[MylesBorins]

I've thrown a baking for lts label on this PR... if the PR shouldn't land on v6.x please remove and add don't land label